WO2024021379A1 - 一种基于无线信号技术的尿布湿度检测方法 - Google Patents

一种基于无线信号技术的尿布湿度检测方法 Download PDF

Info

Publication number
WO2024021379A1
WO2024021379A1 PCT/CN2022/133189 CN2022133189W WO2024021379A1 WO 2024021379 A1 WO2024021379 A1 WO 2024021379A1 CN 2022133189 W CN2022133189 W CN 2022133189W WO 2024021379 A1 WO2024021379 A1 WO 2024021379A1
Authority
WO
WIPO (PCT)
Prior art keywords
diaper
crs
signal
sample set
heuristic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2022/133189
Other languages
English (en)
French (fr)
Inventor
陈艳姣
徐文渊
邓江毅
薛梦
白怡杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University
Original Assignee
Zhejiang University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang University filed Critical Zhejiang University
Priority to US18/278,412 priority Critical patent/US12687502B2/en
Publication of WO2024021379A1 publication Critical patent/WO2024021379A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N22/00—Investigating or analysing materials by the use of microwaves or radio waves, i.e. electromagnetic waves with a wavelength of one millimetre or more
    • G01N22/04—Investigating moisture content
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/06—Systems determining position data of a target
    • G01S13/08—Systems for measuring distance only
    • G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
    • G01S13/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/67—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation

Definitions

  • the present invention relates to the field of intelligent health detection technology, and more specifically to a diaper wetness detection method based on wireless signal technology.
  • Existing diaper wetness detection methods include: using visual aids as wetness indicators, setting Bluetooth sensors on diapers; and installing RFID tags on diapers.
  • the above detection solutions for diaper wetness are actually active inspections of diapers or the installation of specific sensors on diapers, which puts a lot of burden on users. Therefore, how to detect diaper wetness without installing any sensors has been a problem that those skilled in the art have been studying.
  • Wireless radio frequency is a non-contact automatic identification technology. Compared with traditional magnetic card and IC card technology, radio frequency technology has the characteristics of non-contact, fast reading speed, and no wear and tear. Wireless radio frequency technology performs contactless two-way data transmission between the reader and the radio frequency card to achieve the purpose of target identification and data exchange. Compared with traditional barcodes, magnetic cards and IC cards, radio frequency cards are non-contact, fast reading, not affected by the environment, long life, and easy to use. They also have anti-collision functions and can handle multiple cards at the same time.
  • the present invention provides a diaper wetness detection method based on wireless signal technology, which can effectively extract the representation of wet diapers from radio frequency signals, eliminate external mobile and static interference, and train excellent diapers through heuristic migration networks.
  • the model detects diaper wetness and can be well adapted to new users in new families.
  • a diaper wetness detection method based on wireless signal technology includes the following steps:
  • the training sample set and the test sample set are input into the heuristic transfer network for training until the loss function converges, the trained heuristic transfer network is obtained, and the diaper wetness of the test sample set is obtained.
  • the expressions of the transmission signal S T (t) and the return signal S R (t) of the radar device are:
  • f c represents the starting frequency
  • B s represents the scanning bandwidth
  • T s represents the scanning time
  • ⁇ represents the reflection delay of the transmitted signal, and the attenuation of the amplitude is ignored.
  • obtaining continuous radar snapshots includes the following steps:
  • S M2 (t 1 , t 2 ,..., t 64 ) represents the time series of S M2 .
  • the background noise elimination method is used to filter multipath interference by collecting the dot product of the CRS 0 filter template on an empty bed and the continuous radar snapshot signal:
  • CRS' CRS*CRS 0 ;
  • the CRS b filter template of the user lying on the bed without wearing diapers can be collected and the continuous radar snapshot signal is dot producted to detect the noise and the Filter static environmental factors and dynamic environment interference:
  • CRS CRS’*CRS b ;
  • CRS 0 is the CRS of an empty bed
  • CRS b is the CRS of the user lying on the bed without wearing diapers.
  • obtaining the outline of the diaper includes the following steps:
  • the time-frequency representation is refined to obtain the second representation of the diaper outline
  • w l represents the instantaneous frequency
  • ⁇ ( ) represents the Kronecker Dirac function
  • ⁇ is a small threshold, set to 1e-8.
  • obtaining the trained heuristic transfer network includes the following steps:
  • the heuristic transfer network is used to represent the transfer difference between the training sample set and the test sample set, and a generative adversarial network mechanism is used to eliminate the transfer difference until the loss function of the heuristic transfer network converges, and the trained heuristic transfer network is obtained , and obtain the diaper wetness of the test sample set in the feature presentation;
  • the loss function L(Ttm) of the heuristic transfer network is:
  • L Rem represents the loss function to eliminate migration differences
  • L Pro represents the loss function to predict diaper wetness
  • F i represents the extracted features
  • H(xi ) represents the migration difference
  • D(F i ) represents the extracted feature distribution
  • CrossEntropy represents the calculation of the cross entropy of two variables
  • y i represents the label of the sample
  • D N represents the source domain set
  • D A represents the set of target domains.
  • the present invention provides a diaper wetness detection method based on wireless signal technology, which has the following beneficial effects:
  • the present invention detects diapers through frequency modulated continuous wave (FMCW) signals, using Continuous radio snapshots (CRS) extracted from reflected radio signals sense static and dynamic environmental information in the radar field of view (FOV), eliminating irrelevant static reflections and irrelevant ones through a filtering template based on CRS and object breathing moving reflection, and then extract the outline of the diaper through wavelet synchronous compression transformation.
  • the representation of the wet diaper can be extracted from the radio frequency signal, eliminating external mobile and static interference.
  • the heuristic migration network can be well Solve the problem of adapting to new users in new households.
  • Figure 1 is a schematic flow chart of the diaper moisture detection method in the present invention
  • Figure 2 is a continuous wireless snapshot of a dry diaper
  • Figure 3 is a continuous wireless snapshot of a wet diaper
  • Figure 4 is the network structure diagram of the heuristic migration network.
  • an embodiment of the present invention discloses a diaper wetness detection method based on wireless signal technology, as shown in Figure 1, including the following steps:
  • Filter templates based on continuous radar snapshots and object breathing to eliminate interference from static and dynamic environmental factors
  • the solution proposed in this embodiment is passive, ubiquitous, comfortable, non-contact, and non-invasive and does not require the installation of sensors or tags on the surface or inside of the diaper, even if the environment around the diaper is dark. Radio signals can be used to extract the unique contour patterns of different diapers to detect the wetness of the diaper worn by the subject.
  • f c represents the starting frequency
  • B s represents the scanning bandwidth
  • T s represents the scanning time
  • ⁇ represents the reflection delay of the transmitted signal, and the attenuation of the amplitude is ignored.
  • obtaining continuous radar snapshots includes the following steps:
  • Doppler-FFT Doppler-Fast Fourier Transform
  • S M2 (t 1 , t 2 ,..., t 64 ) represents the time series of S M2 .
  • the background noise elimination method is used to filter multipath interference by collecting the dot product of the CRS 0 filter template on an empty bed and the continuous radar snapshot signal:
  • CRS' CRS*CRS 0 ;
  • the CRS b filter template of the user lying on the bed without wearing diapers can be collected and the continuous radar snapshot signal is dot producted to calculate the noise and the Filter static environmental factors and dynamic environment interference:
  • CRS CRS’*CRS b ;
  • CRS 0 is the CRS of an empty bed
  • CRS b is the CRS of the user lying on the bed without wearing diapers.
  • obtaining the outline of the diaper specifically includes the following steps:
  • the time-frequency representation is refined to obtain the second representation of the diaper outline
  • w l represents the instantaneous frequency
  • ⁇ ( ) represents the Kronecker Dirac function
  • ⁇ is a small threshold, set to 1e-8.
  • obtaining the trained heuristic transfer network includes the following steps:
  • the heuristic transfer network is used to represent the transfer difference between the training sample set and the test sample set, and a generative adversarial network mechanism is used to eliminate the transfer difference until the loss function of the heuristic transfer network converges, and the trained heuristic transfer network is obtained, and Obtain the diaper wetness of the test sample set in the feature rendering;
  • the loss function L(Ttm) of the heuristic transfer network is:
  • L Rem represents the loss function to eliminate migration differences
  • L Pro represents the loss function to predict diaper wetness
  • F i represents the extracted features
  • H(xi ) represents the migration difference
  • D(F i ) represents the extracted feature distribution
  • CrossEntropy represents the calculation of the cross entropy of two variables
  • y i represents the label of the sample
  • D N represents the source domain set
  • D A represents the set of target domains.
  • the wet diaper representation can be effectively extracted from the radio frequency signal, eliminating external static interference and dynamic interference, and migrating the network through heuristics
  • the training model solves the problem of using radio signals to detect diaper wetness under different scenarios, distances, directions and other conditions, and completes the detection of diaper wetness without installing sensors.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Pathology (AREA)
  • Medical Informatics (AREA)
  • Artificial Intelligence (AREA)
  • Multimedia (AREA)
  • Immunology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Electromagnetism (AREA)
  • Chemical & Material Sciences (AREA)
  • Computing Systems (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Databases & Information Systems (AREA)
  • Physiology (AREA)
  • Psychiatry (AREA)
  • Signal Processing (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Absorbent Articles And Supports Therefor (AREA)
  • Orthopedics, Nursing, And Contraception (AREA)

Abstract

本发明公开了一种基于无线信号技术的尿布湿度检测方法，涉及智能健康检测技术领域，包括：通过雷达设备向尿布发送调频连续波信号，获取传输信号ST(t)及返回信号SR(t)；检测静态环境因素和动态环境因素，获取连续雷达快照；基于连续雷达快照和物体呼吸的过滤模板，消除环境因素的干扰；利用小波同步压缩变换，获取尿布的轮廓，分为训练样本集和测试样本集；将训练样本集和测试样本集输入启发式迁移网络中进行训练，直至损失函数收敛，得到训练完成的启发式迁移网络，并获取测试样本集的尿布湿度。本发明中的技术方案能够有效的从射频信号中提取湿尿布的表示，消除外部的移动和静态干扰，通过启发式迁移网络解决适应新家庭中的新人的问题。

Description

一种基于无线信号技术的尿布湿度检测方法 技术领域
本发明涉及智能健康检测技术领域，更具体的说是涉及一种基于无线信号技术的尿布湿度检测方法。
背景技术
现有的尿布湿度检测方法包括：使用视觉辅助设备作为湿度指示器，将蓝牙传感器设置在尿布上；在尿布上安装RFID标签。以上关于尿布湿度的检测方案实际上是对尿布的主动检查或者在尿布上安装特定的传感器，给用户带来了许多负担。因此，如何在不安装任何传感器的基础上对尿布湿度进行检测是本领域技术人员一直研究的问题。
无线射频是一种非接触式的自动识别技术，射频技术相对于传统的磁卡及IC卡技术具有非接触、阅读速度快、无磨损等特点。无线射频技术在阅读器和射频卡之间进行非接触双向数据传输，以达到目标识别和数据交换的目的。与传统的条形码、磁卡及IC卡相比，射频卡具有非接触、阅读速度快、不受环境影响、寿命长、便于使用的特点，同时还具有防冲突功能，能同时处理多张卡片。
为了实现一个被动的、舒适的、无处不在的、基于射频的尿片湿度检测方案，还面临着一些问题：1.如何从射频信号中提取湿尿布的表示；2.如何消除外部的移动和静态干扰；3.如何使该方案适应新家庭中的新用户，以上问题是本领域技术人员亟需解决的技术问题。
发明内容
有鉴于此，本发明提供了一种基于无线信号技术的尿布湿度检测方法，可以有效的从射频信号中提取湿尿布的表示，消除外部的移动和静态干扰，通过启发式迁移网络训练出优秀的模型检测尿布湿度，可以很好的适应新家庭中的新用户。
为了实现上述目的，本发明提供如下技术方案：
一种基于无线信号技术的尿布湿度检测方法，包括以下步骤：
通过雷达设备向尿布发送调频连续波信号，获取传输信号S T(t)及所述雷达设备的返回信号S R(t)；
基于所述传输信号S T(t)及返回信号S R(t)，检测静态环境因素和动态环境因素，获取连续雷达快照；
基于所述连续雷达快照和物体呼吸的过滤模板，消除所述静态环境因素和动态环境因素的干扰；
利用小波同步压缩变换，获取尿布的轮廓，分为训练样本集和测试样本集；
将所述训练样本集和测试样本集输入启发式迁移网络中进行训练，直至损失函数收敛，得到训练完成的启发式迁移网络，并获取所述测试样本集的尿布湿度。
可选的，所述传输信号S T(t)及所述雷达设备的返回信号S R(t)的表达式为：
Figure PCTCN2022133189-appb-000001
Figure PCTCN2022133189-appb-000002
其中，f c表示起始频率；B s表示扫描宽带；T s表示扫描时间；τ表示发射信号的反射时延，并忽略了振幅的衰减。
可选的，所述获取连续雷达快照，包括以下步骤：
将所述传输信号S T(t)和返回信号S R(t)相乘并过滤频率最高的部分，获得信号S M1(t)；
Figure PCTCN2022133189-appb-000003
通过距离维快速傅里叶变换检测静态环境因素，获取信号S M2(t)；
S M2(t)＝rangeFFT(S M1(t))；
通过多普勒-快速傅里叶变换检测静态环境因素和动态环境因素，获取连续雷达快照CRS(t)；
CRS(t)＝dopplerFFT(S M2(t 1,t 2,...,t 64))；
其中，S M2(t 1,t 2,...,t 64)表示S M2的时间序列。
可选的，消除所述静态环境因素和动态环境因素的干扰，具体为：
由于多径干扰复杂且难以建模，因此采用背景噪声消除的方法，通过收集一个空床上的CRS 0过滤模板与所述连续雷达快照信号点积，过滤多路径干扰：
CRS'＝CRS*CRS 0；
由于尿布总是穿在身上，个人的呼吸也会包含在CRS中，因此可通过收集用户不穿尿布躺在床上的CRS b过滤模板与所述连续雷达快照信号做点积，对噪音及所述静态环境因素和动态环境的干扰进行过滤：
CRS”＝CRS'*CRS b；
其中，CRS 0是空床的CRS，CRS b为用户不穿尿布躺在床上的CRS。
可选的，所述获取尿布的轮廓，包括以下步骤：
采用小波同步压缩变换，消除尿布时频图上的涂抹现象，获取尿布轮廓的第一表示；
Figure PCTCN2022133189-appb-000004
其中，w b为
Figure PCTCN2022133189-appb-000005
的中心值，τ表示分辨率时域，Δw＝w b-w b-1，a k表示尺度的离散值，A表示
Figure PCTCN2022133189-appb-000006
Figure PCTCN2022133189-appb-000007
f(t)表示湿尿布对应的信号，ψ( )表示小波基函数；(Δa) k＝a k-a k-1；
通过多次小波同步压缩变换，细化时频表示，获取尿布轮廓的第二表示；
Figure PCTCN2022133189-appb-000008
其中，w l表示瞬时频率，δ( )表示克罗内克狄拉克函数，
Figure PCTCN2022133189-appb-000009
表示Wsst [M]剖面的瞬时频率估计；
通过以下约束条件限制M的取值；
Figure PCTCN2022133189-appb-000010
其中，ε是一个很小的阈值，设置为1e-8。
可选的，得到训练完成的启发式迁移网络，包括以下步骤：
将所述训练样本集和测试样本集输入启发式迁移网络中进行训练；
采用Resnet网络提取特征表示；
通过启发式迁移网络表示训练样本集和测试样本集之间的迁移差异，并采用生成式对抗网络机制消除所述迁移差异，直至启发式迁移网络的损失函数收敛，得到训练完成的启发式迁移网络，并在特征呈现中获取测试样本集的尿布湿度；
其中，启发式迁移网络的损失函数L(Ttm)为：
L(Ttm)＝L Rem+L Pro；
Figure PCTCN2022133189-appb-000011
L Pro＝L CrossEntropy(F i-H(x i),y i)；
上式中：L Rem表示消除迁移差异的损失函数，L Pro表示预测尿布湿度的损失函数；F i表示提取的特征，H(x i)表示迁移差异，
Figure PCTCN2022133189-appb-000012
表示源域的期望，D(F i)表示提取的特征分布，
Figure PCTCN2022133189-appb-000013
表示目标域的期望，L CrossEntropy表示计算两个变量的交叉熵，y i表示样本的标签，
Figure PCTCN2022133189-appb-000014
表示来自源域的信号，D N表示源域集合，
Figure PCTCN2022133189-appb-000015
表示来自目标域的信号，D A表示目标域的集合。
经由上述的技术方案可知，与现有技术相比，本发明公开提供了一种基于无线信号技术的尿布湿度检测方法，具有以下有益效果：本发明通过调频连续波(FMCW)信号检测尿布，利用从反射无线电信号中提取的连续无线快照(CRS)感知雷达视场(FOV)中的静态环境信息和动态环境信息，通过一个基于CRS和物体呼吸的过滤模板，消除不相关的静态反射和不相关的移动反射，再通过小波同步压缩变换提取尿布的轮廓，通过本发明的技术方案可以从射频信号中提取湿尿布的表示，消除外部的移动和静态干扰，同时，启发式迁移网络可以很好地解决适应新家庭中的新用户的问题。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据提供的附图获得其他的附图。
图1为本发明中尿布湿度检测方法的流程示意图；
图2为干尿布的连续无线快照；
图3为湿尿布的连续无线快照；
图4为启发式迁移网络的网络结构图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
首先，通过研究无线电信号来捕捉受试者身上尿布湿度的可行性，发现从无线电信号中提取的尿布轮廓对不同的湿尿布呈现出独特的模式。基于此，本发明实施例公开了一种基于无线信号技术的尿布湿度检测方法，如图1所示，包括以下步骤：
通过雷达设备向尿布发送调频连续波信号，获取传输信号S T(t)及雷达设备的返回信号S R(t)；
基于传输信号S T(t)及返回信号S R(t)，检测静态环境因素和动态环境因素，获取连续雷达快照；
基于连续雷达快照和物体呼吸的过滤模板，消除静态环境因素和动态环境因素的干扰；
利用小波同步压缩变换，获取尿布的轮廓，分为训练样本集和测试样本集；
将训练样本集和测试样本集输入启发式迁移网络中进行训练，直至损失函数收敛，得到训练完成的启发式迁移网络，并获取测试样本集的尿布湿度。
本实施例中提出的解决方案是被动的、无所不在的、舒适的、非接触的、非侵入性的，不需要在尿布的表面或内部安装传感器或标签，即使尿布周围 的环境是黑暗的，也可以利用无线电信号提取不同尿布的独特轮廓图案，以检测受试者身上所穿的尿布湿度。
具体方案如下：
向尿布发送调频连续波信号，获取传输信号S T(t)及雷达设备的返回信号S R(t)，其表达式为：
Figure PCTCN2022133189-appb-000016
Figure PCTCN2022133189-appb-000017
其中，f c表示起始频率；B s表示扫描宽带；T s表示扫描时间；τ表示发射信号的反射时延，并忽略了振幅的衰减。
进一步地，获取连续雷达快照，包括以下步骤：
将传输信号S T(t)和返回信号S R(t)相乘并过滤频率最高的部分，获得信号S M1(t)；
Figure PCTCN2022133189-appb-000018
通过距离维快速傅里叶变换(range-FFT)检测静态环境因素，获取信号S M2(t)；
S M2(t)＝rangeFFT(S M1(t))；
通过多普勒-快速傅里叶变换(doppler-FFT)检测静态环境因素和动态环境因素，获取连续雷达快照CRS(t)；
CRS(t)＝dopplerFFT(S M2(t 1,t 2,...,t 64))；
其中，S M2(t 1,t 2,...,t 64)表示S M2的时间序列。
进一步地，消除静态环境因素和动态环境因素的干扰，具体为：
由于多径干扰复杂且难以建模，在本实施例中采用背景噪声消除的方法，通过收集一个空床上的CRS 0过滤模板与所述连续雷达快照信号点积，过滤多路径干扰：
CRS'＝CRS*CRS 0；
由于尿布总是穿在身上，个人的呼吸也会包含在CRS中，因此可以通过收集用户不穿尿布躺在床上的CRS b过滤模板与所述连续雷达快照信号做点积，对噪音及所述静态环境因素和动态环境的干扰进行过滤：
CRS”＝CRS'*CRS b；
其中，CRS 0是空床的CRS，CRS b为用户不穿尿布躺在床上的CRS。
进一步地，获取尿布的轮廓(参见图2、图3)，具体包括以下步骤：
采用小波同步压缩变换，消除尿布时频图上的涂抹现象，获取尿布轮廓的第一表示；
Figure PCTCN2022133189-appb-000019
其中，w b为
Figure PCTCN2022133189-appb-000020
的中心值，τ表示分辨率时域，Δw＝w b-w b-1，a k表示尺度的离散值，A表示
Figure PCTCN2022133189-appb-000021
Figure PCTCN2022133189-appb-000022
f(t)表示湿尿布对应的信号，ψ( )表示小波基函数；(Δa) k＝a k-a k-1；
通过多次小波同步压缩变换，细化时频表示，获取尿布轮廓的第二表示；
Figure PCTCN2022133189-appb-000023
其中，w l表示瞬时频率，δ( )表示克罗内克狄拉克函数，
Figure PCTCN2022133189-appb-000024
表示Wsst [M]剖面的瞬时频率估计；
通过以下约束条件限制M的取值；
Figure PCTCN2022133189-appb-000025
其中，ε是一个很小的阈值，设置为1e-8。
进一步地，得到训练完成的启发式迁移网络，包括以下步骤：
将训练样本集和测试样本集输入启发式迁移网络(参见图4)中进行训练；
采用Resnet网络提取特征表示；
通过启发式迁移网络表示训练样本集和测试样本集之间的迁移差异，并采用生成式对抗网络机制消除迁移差异，直至启发式迁移网络的损失函数收敛，得到训练完成的启发式迁移网络，并在特征呈现中获取测试样本集的尿布湿度；
其中，启发式迁移网络的损失函数L(Ttm)为：
L(Ttm)＝L Rem+L Pro；
Figure PCTCN2022133189-appb-000026
L Pro＝L CrossEntropy(F i-H(x i),y i)；
上式中：L Rem表示消除迁移差异的损失函数，L Pro表示预测尿布湿度的损失函数；F i表示提取的特征，H(x i)表示迁移差异，
Figure PCTCN2022133189-appb-000027
表示源域的期望，D(F i)表示提取的特征分布，
Figure PCTCN2022133189-appb-000028
表示目标域的期望，L CrossEntropy表示计算两个变量的交叉熵，y i表示样本的标签，
Figure PCTCN2022133189-appb-000029
表示来自源域的信号，D N表示源域集合，
Figure PCTCN2022133189-appb-000030
表示来自目标域的信号，D A表示目标域的集合。
传统的尿布湿度检测方法会给用户带来很大负担，而基于本发明中的技术方案，可以有效的从射频信号中提取湿尿布表示，消除外部的静态干扰和动态干扰，通过启发式迁移网络训练模型解决了使用无线电信号在不同场景、 距离、方向和其他条件下检测尿布湿度的难题，在不安装传感器的前提下，完成对尿布湿度的检测。
对所公开的实施例的上述说明，使本领域专业技术人员能够实现或使用本发明。对这些实施例的多种修改对本领域的专业技术人员来说将是显而易见的，本文中所定义的一般原理可以在不脱离本发明的精神或范围的情况下，在其它实施例中实现。因此，本发明将不会被限制于本文所示的这些实施例，而是要符合与本文所公开的原理和新颖特点相一致的最宽的范围。

Claims (6)

  1. 一种基于无线信号技术的尿布湿度检测方法，其特征在于，包括以下步骤：
    通过雷达设备向尿布发送调频连续波信号，获取传输信号S T(t)及所述雷达设备的返回信号S R(t)；
    基于所述传输信号S T(t)及返回信号S R(t)，检测静态环境因素和动态环境因素，获取连续雷达快照；
    基于所述连续雷达快照和物体呼吸的过滤模板，消除所述静态环境因素和动态环境因素的干扰；
    利用小波同步压缩变换，获取尿布的轮廓，分为训练样本集和测试样本集；
    将所述训练样本集和测试样本集输入启发式迁移网络中进行训练，直至损失函数收敛，得到训练完成的启发式迁移网络，并获取所述测试样本集的尿布湿度。
  2. 根据权利要求1所述的一种基于无线信号技术的尿布湿度检测方法，其特征在于，所述传输信号S T(t)及所述雷达设备的返回信号S R(t)的表达式为：
    Figure PCTCN2022133189-appb-100001
    Figure PCTCN2022133189-appb-100002
    其中，f c表示起始频率；B s表示扫描宽带；T s表示扫描时间；τ表示发射信号的反射时延，并忽略了振幅的衰减。
  3. 根据权利要求2所述的一种基于无线信号技术的尿布湿度检测方法，其特征在于，所述获取连续雷达快照，包括以下步骤：
    将所述传输信号S T(t)和返回信号S R(t)相乘并过滤频率最高的部分，获得信号S M1(t)；
    Figure PCTCN2022133189-appb-100003
    通过距离维快速傅里叶变换检测静态环境因素，获取信号S M2(t)；
    S M2(t)＝rangeFFT(S M1(t))；
    通过多普勒-快速傅里叶变换检测静态环境因素和动态环境因素，获取连续雷达快照CRS(t)；
    CRS(t)＝dopplerFFT(S M2(t 1,t 2,...,t 64))；
    其中，S M2(t 1,t 2,...,t 64)表示S M2的时间序列。
  4. 根据权利要求1所述的一种基于无线信号技术的尿布湿度检测方法，其特征在于，消除所述静态环境因素和动态环境因素的干扰，具体为：
    采用背景噪声消除的方法，通过收集一个空床上的CRS 0过滤模板与所述连续雷达快照信号点积，过滤多路径干扰：
    CRS'＝CRS*CRS 0；
    通过收集用户不穿尿布躺在床上的CRS b过滤模板与所述连续雷达快照信号做点积，对噪音及所述静态环境因素和动态环境的干扰进行过滤：
    CRS”＝CRS'*CRS b；
    其中，CRS 0是空床的CRS，CRS b为用户不穿尿布躺在床上的CRS。
  5. 根据权利要求1所述的一种基于无线信号技术的尿布湿度检测方法，其特征在于，所述获取尿布的轮廓，包括以下步骤：
    采用小波同步压缩变换，消除尿布时频图上的涂抹现象，获取尿布轮廓的第一表示；
    Figure PCTCN2022133189-appb-100004
    其中，w b为
    Figure PCTCN2022133189-appb-100005
    的中心值，τ表示分辨率时域，Δw＝w b-w b-1，a k表示尺度的离散值，A表示
    Figure PCTCN2022133189-appb-100006
    Figure PCTCN2022133189-appb-100007
    f(t)表示湿尿布对应的信号，ψ()表示小波基函数；(Δa) k＝a k-a k-1；
    通过多次小波同步压缩变换，细化时频表示，获取尿布轮廓的第二表示；
    Figure PCTCN2022133189-appb-100008
    其中，w l表示瞬时频率，δ()表示克罗内克狄拉克函数，
    Figure PCTCN2022133189-appb-100009
    表示Wsst [M]剖面的瞬时频率估计；
    通过以下约束条件限制M的取值；
    Figure PCTCN2022133189-appb-100010
    其中，ε是一个很小的阈值，设置为1e-8。
  6. 根据权利要求1所述的一种基于无线信号技术的尿布湿度检测方法，其特征在于，得到训练完成的启发式迁移网络，包括以下步骤：
    将所述训练样本集和测试样本集输入启发式迁移网络中进行训练；
    采用Resnet网络提取特征表示；
    通过启发式迁移网络表示训练样本集和测试样本集之间的迁移差异，并采用生成式对抗网络机制消除所述迁移差异，直至启发式迁移网络的损失函数收敛，得到训练完成的启发式迁移网络，并在特征呈现中获取测试样本集的尿布湿度；
    其中，启发式迁移网络的损失函数L(Ttm)为：
    L(Ttm)＝L Rem+L Pro；
    Figure PCTCN2022133189-appb-100011
    L Pro＝L CrossEntropy(F i-H(x i),y i)；
    上式中：L Rem表示消除迁移差异的损失函数，L Pro表示预测尿布湿度的损失函数；F i表示提取的特征，H(x i)表示迁移差异，
    Figure PCTCN2022133189-appb-100012
    表示源域的期望，D(F i)表示提取的特征分布，
    Figure PCTCN2022133189-appb-100013
    表示目标域的期望，L CrossEntropy表示计算两个变量的交叉熵，y i表示样本的标签，
    Figure PCTCN2022133189-appb-100014
    表示来自源域的信号，D N表示源域集合，
    Figure PCTCN2022133189-appb-100015
    表示来自目标域的信号，D A表示目标域的集合。
PCT/CN2022/133189 2022-07-29 2022-11-21 一种基于无线信号技术的尿布湿度检测方法 Ceased WO2024021379A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/278,412 US12687502B2 (en) 2022-07-29 2022-11-21 Method for detecting diaper wetness based on radio signal technology

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210904762.9A CN115187815A (zh) 2022-07-29 2022-07-29 一种基于无线信号技术的尿布湿度检测方法
CN202210904762.9 2022-07-29

Publications (1)

Publication Number Publication Date
WO2024021379A1 true WO2024021379A1 (zh) 2024-02-01

Family

ID=83520789

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/133189 Ceased WO2024021379A1 (zh) 2022-07-29 2022-11-21 一种基于无线信号技术的尿布湿度检测方法

Country Status (3)

Country Link
US (1) US12687502B2 (zh)
CN (1) CN115187815A (zh)
WO (1) WO2024021379A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115187815A (zh) 2022-07-29 2022-10-14 浙江大学 一种基于无线信号技术的尿布湿度检测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2006119523A1 (de) * 2005-05-09 2006-11-16 Diaplan Elektronic Gmbh Krankenpflegesystem
CN110361071A (zh) * 2018-04-11 2019-10-22 英飞凌科技股份有限公司 使用毫米波雷达传感器进行液体检测的装置和方法
CN111461269A (zh) * 2020-04-07 2020-07-28 南京大学 基于射频识别技术的尿不湿湿度检测系统及其工作方法
CN114184645A (zh) * 2021-12-30 2022-03-15 南京楚航科技有限公司 一种基于毫米波雷达的尿不湿潮湿检测告警设备
CN115187815A (zh) * 2022-07-29 2022-10-14 浙江大学 一种基于无线信号技术的尿布湿度检测方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5861179B1 (ja) * 2015-06-04 2016-02-16 株式会社Ｃｑ−Ｓネット 定在波レーダーによる状態検知装置
US10548476B2 (en) * 2017-08-17 2020-02-04 Welch Allyn, Inc. Patient monitoring system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2006119523A1 (de) * 2005-05-09 2006-11-16 Diaplan Elektronic Gmbh Krankenpflegesystem
CN110361071A (zh) * 2018-04-11 2019-10-22 英飞凌科技股份有限公司 使用毫米波雷达传感器进行液体检测的装置和方法
CN111461269A (zh) * 2020-04-07 2020-07-28 南京大学 基于射频识别技术的尿不湿湿度检测系统及其工作方法
CN114184645A (zh) * 2021-12-30 2022-03-15 南京楚航科技有限公司 一种基于毫米波雷达的尿不湿潮湿检测告警设备
CN115187815A (zh) * 2022-07-29 2022-10-14 浙江大学 一种基于无线信号技术的尿布湿度检测方法

Also Published As

Publication number Publication date
US12687502B2 (en) 2026-07-21
US20250012738A1 (en) 2025-01-09
CN115187815A (zh) 2022-10-14

Similar Documents

Publication Publication Date Title
CN112716474B (zh) 基于生物微波雷达的非接触式睡眠状态监测方法和系统
CN112401856B (zh) 一种基于毫米波雷达的疗养院监护方法及系统
CN111759305B (zh) 活体检测方法与活体检测系统
Huang et al. PhaseAnti: An anti-interference WiFi-based activity recognition system using interference-independent phase component
JP2021099348A (ja) センサを用いたジェスチャ認識
US20220318548A1 (en) Method, apparatus, computing device and computer-readable storage medium for identifying signal
CN109444813A (zh) 一种基于bp和dnn双神经网络的rfid室内定位方法
CN109512390B (zh) 基于eeg时域多维度特征及m-wsvm的睡眠分期方法及可穿戴装置
Rivera et al. Radar-based fall detection exploiting time-frequency features
CN114465678A (zh) 一种基于深度学习的复杂活动wifi感知方法
CN111990968A (zh) 一种信号监测方法、设备、存储介质及系统
Venkatnarayan et al. WiFi based multi-user gesture recognition
CN116359916A (zh) 一种基于毫米波雷达传感器的人员失禁检测方法
CN115187815A (zh) 一种基于无线信号技术的尿布湿度检测方法
Moghaddam et al. Device-free human activity recognition: A systematic literature review
Fusco et al. Deep learning classifier for robust artifact rejection in FMCW radar vital sensing
Xue et al. Wet-Ra: Monitoring diapers wetness with wireless signals
CN105962946A (zh) 一种非接触式的人体睡姿安全检测方法及系统
CN116869516A (zh) 一种基于多源异构数据的综合运动评估方法及系统
Yang et al. RFID-based driving fatigue detection
Golipoor et al. Environment and person-independent gesture recognition with non-static RFID tags leveraging adaptive signal segmentation
CN113919402A (zh) 基于rfid的密封液体识别系统
Lopez et al. Walk and jog characterization using a triaxial accelerometer
CN109316174A (zh) 一种基于波形设计与反射波处理的生命体征无线监测方法
Xue et al. Meta-learning for human-centered wireless sensing: architecture, applications, and challenges

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 18278412

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22952814

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22952814

Country of ref document: EP

Kind code of ref document: A1

WWG Wipo information: grant in national office

Ref document number: 18278412

Country of ref document: US